Advertisement

Classification of sites according to gamma radiation levels in ambient air using GIS and data mining techniques

  • Mazhar Mahmoud Hefnawi
  • Ahmed Yousef El-Haseib
Original Paper
  • 106 Downloads

Abstract

Purpose

This work uses the data which represent the measurements of the gamma radiation levels in ambient air from many gamma monitoring stations that are distributed in many sites to classify the regions which cover these sites according to these measurements.

Method

The processes of the classification are: dividing the range of measurements to several intervals, making interpolation for all regions that cover all gamma monitoring stations, representing the interpolation information in the map using geographic information systems technology and finally classifying all the sites on this map according to the determined intervals by using data mining techniques via interpolation information.

Implementation and Importance

This method is implemented for determining the background of gamma radiation levels for many sites in Egypt. This background is necessary for many environmental researches because it is useful for making risk assessment evaluation for any site in Egypt.

Results

The output result from this implementation shows that most sites in Egypt have been classified within three intervals: first interval is from 2.02E−2 to 4.75E−2μSv/h with 47.11% of Egypt area, second interval is from 4.75E−2 to 8.85E−2μSv/h with 40% of Egypt area and third interval is from 8.85E−2 to 1.42E−1μSv/h with 6.82% of Egypt area.

Conclusions

This method is more useful than other traditional methods because the results from this method show that this method saves more effort, time and cost than other methods.

Keywords

GIS Data mining Environmental classification Data modeling 

Introduction

Classifying streams is very necessary subject that must be received enough attention in the scientific literature. In particular, it is very important to solve the problem for specific needs for the field based. When referring to streams which represent measurements of environmental data in a continuous way such metrological data, gamma radiation data in ambient air and seasonal data several ways are used to define perennial, intermittent and ephemeral [1]. This study uses definitions from the USGS [2]. The definitions are related to time and state that a perennial stream is one which flows continuously; an intermittent (or seasonal) stream is one which flows only at certain times of the year when it receives water from springs or from some surface source such as melting snow in mountainous areas; and an ephemeral stream is one that flows only in direct response to precipitation, and whose channel is at all times above the water table. It is important to notice that the previous definitions have implications such that a connection of the flow conditions with the water, which in turn is influenced by climate changes and seasonal variations. Geographic information system (GIS) technologies have made an excellent progress in the last several years, making it possible to integrate several data layers for such purposes as mapping and modeling efforts. One data layer that has not been available in a reliable way, especially in a more detailed scale (county and city level, for example), is that of stream classification. There is therefore a need to investigate new approaches and strategies to classify streams that do not depend strongly on field methods. Field methods produce reasonable data if applied properly [3] but can be expensive and are prone to errors due to incorrect measurements or inability to reach certain areas. This study focuses on strategies based on GIS and data mining, using digital information which is the measurements of the gamma radiation levels in ambient air from many gamma monitoring stations, with a case study for many sites in Egypt.

A continuous stream from measurements of the gamma radiation levels in ambient air for any locations is considered as a measure of the ionizing radiation that presents in the environment at a particular location which has many sources. These sources are cosmic radiation and environmental radioactivity from such naturally occurring radioactive materials including radon and radium, and man-made fallout from nuclear weapons testing and nuclear accidents [4].

Background radiation is defined by the International Atomic Energy Agency as “dose or dose rate (or an observed measure related to the dose or dose rate)” to all sources other than the one or more specified [5]. So there should be a difference between dose which is already in a location, which is defined here as being “background,” and the dose that is caused by specified source. This is necessary because radiation measurements are taken from a specified radiation source and the existing background may affect this measurement at the same time. Measurement of radioactive contamination in a gamma radiation without knowing the real value of background, could lead to incorrect reading value of the contamination.

If radiation source is not present as being of concern, then the total radiation dose measurement at a location is generally called the background radiation, and this is usually the case where an ambient dose rate is measured for environmental purposes.

It is important to classify the sites according to gamma radiation levels because higher concentrations of gamma radiation may lead to the following events.

Sufficient energy from gamma radiation causes chemical changes in human cells and damages them. Some cells may die or become abnormal, either temporarily or permanently. Damaging the genetic material (DNA) contained in the human’s cells, radiation can cause cancer. Our bodies are extremely efficient at repairing cell damage. The extent of the damage to the cells depends on the amount and duration of the exposure.

There are two types of health effects: chronic (long term) and acute (short term).

Chronic exposure is continuous or intermittent exposure to radiation over a long period of time. With chronic exposure, it may lead to health effect. These effects can include cancer and other health outcomes such as benign tumors, cataracts and potentially harmful genetic changes.

Acute health effects occur when large parts of the body are exposed to a large amount of radiation. The large exposure can occur all at once or from multiple exposures in a short period of time. Environmental sources for acute effects are very rare. Nearly all sources of acute exposure are man-made and any type of nuclear accident.

Therefore, it is necessary to classify the sites according to gamma radiation levels to make a good tool for putting a baseline data. These baseline data will be a useful tool to evaluate the processes of risk assessment to avoid acute health effects and decrease the chronic health effects from exposure to sources of gamma radiation.

Materials and methods

The aim of this work is modeling the information of gamma radiation levels for many sites in Egypt and classifying these sites according to the gamma radiation levels. Therefore, this work uses the data which represent the measurements of the gamma radiation levels in ambient air from many gamma monitoring stations that are distributed in many sites to classify the regions which cover these sites according to these measurements. The processes of the classification are as follows [6]:
  1. 1.

    Dividing the range of measurements to several intervals.

     
The process of dividing the range of measurements which uses Natural Breaks method [7], also called the Jenks natural breaks classification method, is a data clustering method designed to determine the best arrangement of values into different classes. This is performed by seeking to minimize each class’s average deviation from the class mean, while maximizing each class’s deviation from the means of the other groups. In other words, the method seeks to reduce the variance within classes and maximize the variance between classes. By using the data from Table 1 as input data to this method, the output intervals after applying Natural Breaks method are listed in Table 2.
Table 1

The list for all gamma stations including the name of the stations and the average measurements of gamma radiation levels in unit of μSv/h for the year 2012

Station number

Station name

Average value for gamma radiation levels at 2012

1

Cairo

3.50E−02

2

Alexandria

2.63E−02

3

Domyat

2.02E−02

4

Port Said

2.68E−02

5

Ismailia

3.50E−02

6

Suez

4.82E−02

7

Hurghada

5.61E−02

8

Aswan

4.17E−02

9

El Dabaa

5.00E−02

10

Mersa Matruh

4.12E−02

11

El Saloom

2.51E−02

12

Tanta

3.61E−02

13

Zagaziq

2.72E−02

14

El Mansoura

3.58E−02

15

Al Arish

2.70E−02

16

Rafah

3.15E−02

17

El Tor

7.36E−02

18

Sharm El Sheikh

1.09E−01

19

Nuweiba

1.09E+00

20

Taba

6.38E−02

21

El Kontella

5.5E−02

22

El Qosema

5.06E−02

23

Saint Catherine

1.34E−01

24

El Fayoum

3.14E−02

25

El kharga

3.55E−02

26

Beni Suef

1.89E−01

27

Asyut

5.37E−02

28

Qena

4.32E−02

29

Al Minya

3.57E−02

Table 2

List of the output intervals after applying Natural Breaks method

Interval number

From

To

Unit

1

2.02E−02

4.75E−02

μSv/h

2

4.75E−02

8.85E−02

μSv/h

3

8.85E−02

1.43E−01

μSv/h

4

1.43E−01

2.53E−01

μSv/h

5

2.53E−01

4.59E−01

μSv/h

6

4.59E−01

7.36E−01

μSv/h

7

7.36E−01

1.09E+00

μSv/h

  1. 2.

    Making interpolation for all regions that cover all gamma monitoring stations.

     
Interpolation is a method of constructing new data points within the range of a discrete set of known data points [8, 9]. In this study, the inverse distance-weighted interpolation method (IDW) is used for constructing new data points within the range for all regions that cover the points which represents gamma monitoring stations. After applying IDW method, nearly all regions, on the map, are covered. Figure 1 shows a distributed gamma monitoring stations in the map of Egypt.
Fig. 1

A distributed gamma monitoring stations in the map of Egypt

  1. 3.

    Representing the interpolation information in the map using GIS technology.

     
This process uses the output result from the interpolation process and represents this output in the map which is shown in Fig. 1. Figure 2 shows the map in Fig. 1 after this process.
Fig. 2

Representation of the interpolation information for the map of Egypt using GIS technology

  1. 4.

    Classifying all the sites on this map according to the determined intervals by using data mining techniques via interpolation information.

     
Using data mining techniques in this process [10] means that an analysis will be performed for the result from the previous process. In other words, an analysis will be performed in the interpolation information to obtain the classification for many sites in Egypt according to gamma Radiation levels. The target of the classification process is producing many maps from the map in Fig. 2 so that every produced map shows only the distribution of only one interval from Table 2. This can be done by performing the logical operation of mathematical formula to transfer the source imagery raster (layer of interpolation information that appears in Fig. 3) to another imagery raster as shown in Fig. 3.
Fig. 3

Performing the logical operation of mathematical formula to transfer the source imagery raster to another imagery raster

The logical operation of mathematical formula to compute the output raster that contains only all sites that belong to the first interval from Table 2 is [11]:
$$ \left( {\left[ {2012\,{\text{Average}}} \right] \ge 0.00202} \right)\;{\text{and}}\;\left( {\left[ {2012\,{\text{Average}}} \right] < 0.0475} \right) $$
where
  • “2012 Average” is the name of the raster which is appearing in Fig. 1.

  • “0.00202” is the first value in the first interval in Table 2 in unit of μSv/h.

  • “0.0475” is the last value in the first interval in Table 2 in unit of μSv/h.

The output from this formula is only one value of Boolean data type. “In other words, the output will be true value or false value,” because the required output is raster which means number of cells and every cell has only one value. This formula will be repeated to the number of times equal to the number of cells in raster ([2012 Average]).

Figure 3 shows how to extract the required layers from one raster.

Assuming this formula is implemented to the cell that has gamma radiation measurements level of 3.00E−2 μSv/h.
  • The first part of this formula which is ([2012 Average] ≥ 0.00202) will be accomplished as follows

  • [2012 Average] will take value of 3.00E−2 μSv/h; therefore, the first part will become

  • (0.03 ≥ 0.00202) and hence the output result from this part is true

  • The third part will become (0.03 < 0.00475), and hence, the output result from this part is true.

  • Now the second part can be implemented after becoming in the form of true and true. So the final output from this formula is true.

By repeating this formula [12] for every cell in raster ([2012 Average]), we will obtain a new raster that shows all the sites that belong to the first interval in Table 2. Therefore, a new map is produced for showing output raster as in Fig. 4.
Fig. 4

Raster that shows all the sites that belong to the first interval in Table 2

Now from Fig. 4, it is easy to determine all the sites which are belonging to the first interval. These sites are covered with gray color in Fig. 4. By repeating this method with other intervals in Table 2, we can obtain the required classification as in the next section.

Results and discussion

The output result after implementing the specified method from the previous section is shown in Fig. 5. From Fig. 5, we can classify many sites in Egypt according to gamma radiation levels. Table 3 shows all the sites that belong to each interval in Table 2 and their area.
Fig. 5

All the sites that belong to each interval in Table 2

Table 3

All sites that belong to each interval in Table 1 and their area

Interval number

From

To

Sites which belong to interval

Percentage of total area for all sites

Unit

1

2.02E−02

4.75E−02

North Sinai, Suez Channel, Nile Delta, Port Said, Ismailia, Suez, Alexandria, Borg El Arab, El Alameen, Mersa Matruh, Seedy Branee, Sewaa oasis, El Saloom, Part of El Frfra, El Fayoum, Al Minya, Deshna, Nga Hmmad, Koos, Luxor, Bna, Koom Embo, Aswan, El Khraga, Elwhaat Elkhargyaa, Elwhaat Eldakhelea, Paris.

47.11

μSv/h

2

4.75E−02

8.85E−02

El Qoseama, El Qontella, El Tor, El Dabaa, Glala mountain, Monkhfad El Qtara, El whaat El Bahria, Asyut, Sohaag, El Frfra oasis, El Quseir, Hurghada.

41.00

μSv/h

3

8.85E−02

1.43E−01

Ras Gharib, Sharm El Sheikh, Safaga, El Qosyear, El Shayeb mountain, Saint Catherine, Taba.

6.82

μSv/h

4

1.43E−01

2.53E−01

Middle of Sinai and Part of Suez golf.

3.75

μSv/h

5

2.53E−01

4.59E−01

Outside of Nuweiba city and part of El Aqaba golf.

0.69

μSv/h

6

4.59E−01

7.36E−01

Surrounding of Nuweiba city

0.36

μSv/h

7

7.36E−01

1.088

Nuweiba city

0.27

μSv/h

From Table 3, we can explore the following results:
  1. 1.

    About 88.11% of Egypt’s area is belonging to interval one and interval two. In other words “About 88.11% of Egypt's area its minimum gamma radiation level is 2.02E−2 μSv/h and its maximum gamma radiation level is 8.852E−2 μSv/h”.

     
  2. 2.

    The maximum gamma radiation levels in which its value is 1.088 μSv/h are found in Nuweiba city and its area represents about 0.27% of Egypt.

     
  3. 3.

    The middle and south of Sinai have higher concentration of gamma radiation levels than other lands in Egypt.

     

This case study proves that GIS and data mining techniques are more useful than using famous or traditional methods for classifying perennial stream such as decision tree (DT), k-nearest neighbor (k-NN), logistic regression (LogR), naïve Bayes (NB), C4.5, support vector machine (SVM), artificial neural network (ANN) and linear classifier (LC) [13, 14] because all of these methods require a huge number of datasets for obtaining a good result. In this case study, the size of gamma radiation measurements which represent the size of dataset is limited, and therefore, if any of the traditional methods are used, then the quality of the output results will be far from good.

Conclusions

GIS and data mining techniques which are used in this case study have extra features that are not exited in traditional methods which are as follows:
  1. 1.

    This method has the ability to estimate the values of any stream for all points that have not any measured value with their global positioning system (GPS).

     
  2. 2.

    This method has the ability to generate an automated map to show the information of the stream values in a map.

     
  3. 3.

    Data mining techniques as shown from this case study can be implemented with this method in a easy way. With this feature, a lot of knowledge can be discovered from some data.

     

Using GIS and data mining techniques for classifying many sites according to any kind of stream saves a lot of effort, time and cost. This method can be used in the processes of the evaluation of risk assessment and also making references for sitting any region according to continuous stream from measurements of the gamma radiation levels in ambient air or any kind of continuous stream.

References

  1. 1.
    B.M. Gritzner. Stream Flow Definitions: Perennial, Intermittent, and Ephemeral Streams. Department of Geography, South Dakota State University, Draft Report (2003)Google Scholar
  2. 2.
    Gritzner. USGS National Mapping Program Standards for Stream/River Features. Department of Geography, South Dakota State University, Draft Report (2003)Google Scholar
  3. 3.
    B.M. Gritzner. Approaches to Stream Delineation, Classification, and Changes in Stream Properties. Department of Geography, South Dakota State University, Draft Report (2003)Google Scholar
  4. 4.
    P. Sharma, P.K. Meher, K.P. Mishra, J. Radiat. Res. Appl. Sci. 7(4), 595 (2014)CrossRefGoogle Scholar
  5. 5.
    IAEA, Safety Glossary. Terminology used in nuclear safety and radiation protection, 2007 edn. (IAEA, Vienna, 2007)Google Scholar
  6. 6.
    D.J. Peuquet, N. Duan, An event-based spatiotemporal data model (ESTDM) for temporal analysis of geographical data. Int. J. Geogr. Inf. Syst. 9(1), 7 (1995)CrossRefGoogle Scholar
  7. 7.
    B. Jiang, Head/tail breaks: a new classification scheme for data with a heavy-tailed distribution. Prof. Geogr. 65(3), 482 (2013)CrossRefGoogle Scholar
  8. 8.
    Jöran Bergh, Jörgen Löfström, Interpolation Spaces: An Introduction, vol. 223 (Springer, Berlin, 1976)zbMATHGoogle Scholar
  9. 9.
    A. Kufner. Triebel, H., Interpolation Theory, Function Spaces, Differential Operators. Berlin, VEB Deutscher Verlag der Wissenschaften 1978. 528 S., M 87, 50. ZAMM J. Appl. Math. Mechanics/Zeitschrift für ewandte Mathematik und Mechanik 59(12), 756–757 (1979)ADSCrossRefGoogle Scholar
  10. 10.
    S.E. Spielman, J.-C. Thill, Comput. Environ. Urban Syst. 32(2), 110 (2008)CrossRefGoogle Scholar
  11. 11.
    H.J. Miller, J. Han (eds.), Geographic Data Mining and Knowledge Discovery (CRC Press, Boca Raton, 2009)Google Scholar
  12. 12.
    Deren Li, Shuliang Wang, Adv. Spatio-Temporal Anal. 5, p173 (2007)Google Scholar
  13. 13.
    R. Entezari-Maleki, A. Rezaei, B. Minaei-Bidgoli, J. Converg. Inf. Technol. 4(3), 94 (2009)Google Scholar
  14. 14.
    W.H. Chen, S.H. Hsu, H.P. Shen, J. Comput. Oper. Res. 32, 2617 (2005)CrossRefGoogle Scholar

Copyright information

© Institute of High Energy Physics, Chinese Academy of Sciences; Nuclear Electronics and Nuclear Detection Society and Springer Nature Singapore Pte Ltd. 2018

Authors and Affiliations

  • Mazhar Mahmoud Hefnawi
    • 2
  • Ahmed Yousef El-Haseib
    • 1
  1. 1.Radiation Safety DepartmentEgyptian Nuclear and Radiological Regulatory Authority (ENRRA)CairoEgypt
  2. 2.Sitting and Environment DepartmentEgyptian Nuclear and Radiological Regulatory Authority (ENRRA)CairoEgypt

Personalised recommendations